There are many applications of ion beam technology which suffer limitations on focusability due to beam emittance and space charge effects. For example, conventional ion lithography and implantation techniques consist of the radio frequency (RF) acceleration of an ion beam and then focusing the ion beam to a small focal point using optics. The focused beam is then rastered across the implantation sample in a particular pattern. The achievable spatial resolution depends on the focal spot size, which depends on the focusing optical elements (magnetic and electrostatic lenses) and on the ion beam current due to repulsive space charge forces in the beam. High spatial resolution (small focal spot size) requires reduced current. In turn, the reduced current lengthens the exposure time. Writing a pattern with the ion beam requires rastering the ion beam across the sample.
In some techniques, rather than rastering an optically focused ion beam, the ion beam is patterned into a desired pattern by applying a mask in the path of the ion beam. The mask simply blocks portions of the ion beam allowing the patterned portion to be transmitted to the sample. One example using a mask structure in which the expanding plasma induces electric fields in the mask dielectric material to micro-focus the beam is described in Ruhl, et al., “Probing of Electromagnetic Field with Laser Generated Proton Beams”, Deutsches Patent und Markenamt Berlin, Aktenzeichen: 101 48 613.8 (2001), which is incorporated herein by reference.
The laser acceleration of ions is well known in the art and includes extensive research on laser ablation. Such techniques involve the expansion of hot plasma into a vacuum, as a result of laser-irradiation of a surface, and the coupling of the recoil momentum to drive implosions. This plasma expansion is from the front, laser-irradiated surface of the target, which is not suitable for high quality ion beam production. Such methods relying on laser ablation are conventionally used in laser machining and material modification, and are not well suited to lithography techniques.
As illustrated in FIG. 1 and as indicated in recent research, short-pulse laser irradiation of thin foils produces extremely low transverse beam emittance, and an essentially completely charge-neutralized beam from the non-irradiated side of the thin foils. This laser acceleration mechanism is related to the so-called Target Normal Sheath acceleration process first identified by Wilks et at., “Energetic Proton Generation in Ultra-Intense Laser-solid Interactions”, Phys. Plasmas, Vol. 8, pg. 542 (2001), and Hatchett et al., Phys. Plasmas, Vol. 7, pg. 2076 (2000), both of which are incorporated herein by reference. As illustrated in FIG. 1, short pulse laser-irradiation 102 is directed on a target 104, e.g., a thin metallic foil. This irradiation 102 produces an ablation plasma 106 expanding from the front or irradiated surface 108 of the target 104. Also generated is a high-density (e.g., 1019-1020 cm3) of hot, MeV electrons 109, which penetrate the target 104, enveloping the target 104 and producing an electron plasma sheath 110 on the rear, non-irradiated surface 112 of the target 104. The electric field from the charge separation in this sheath 110 is on the order of the hot electron temperature (MeV), divided by the spatial scale length of the sheath (˜micrometer), e.g., equal to approximately 1012 V/m (TV/m). This is sufficient to field ionize atoms in a thin layer on the rear surface 112 of the target 104, either protons from the bulk material 104, contaminants (such as hydrocarbons or water vapor), or purposefully deposited ion source material 114 (e.g., CaF). This rear-surface collimated emission 116 (i.e., the ionized atoms on the non-irradiated surface 112 of the target 104) does not suffer the space charge limitations of conventional non-neutralized beams. Thus, very high currents and current densities can be produced, much in excess of conventional RE acceleration techniques, while producing a proton beam quality having an emittance comparable to conventional RF accelerators.
Following ionization, the positive ions or protons (e.g., illustrated as F7+ ions) are accelerated by the charge separation field, are analogous to a virtual cathode effect (see FIG. 2A which illustrates “Phase I” following ionization). FIG. 2A is a density vs. distance plot, where the distance z extends along an axis normal to the target surface and increasing along a direction of the expansion of the proton and ion beam. It is seen that due to the hot electron emission indicated as the curved portion of the hot electron density curve nhot, positive charges accumulate at the boundary of the target (indicated as the positive charges on the right side of the ion (and proton) density curve nion).
As the ion density scale length increases (by expansion from the initially solid surface) to the order of the hot electron scale length, the ions (e.g., F7+ ions) co-mingle with the hot electrons (e.g., e−) and become a quasineutral plasma (FIG. 2B, “Phase II”). The quasineutral plasma including ions and protons as well as electrons continues to accelerate outward from the rear non-irradiated surface of the target 104 due to the ambipolar electric field. The commingling of ions/protons and electrons is illustrated at region 202 as the region to the right of the ncold curve and shared by the nhot and nion curves.
The quasineutral plasma expansion accelerates the ions and protons to energies of several times the hot electron temperature, i.e., many MeV. Up to 50 MeV protons, and 100 MeV Fluorine ions have been observed in various experiments. The quasineutral expansion can be extremely laminar, and leads to a smoothly expanding plasma. The radial expansion is determined by the initial spatial distribution of the hot electron sheath 110 on the rear surface of the target 104. It is customary to assume Boltzmann equilibrium for the distribution of the hot electrons (although this may not strictly be correct for the very early, few femtosecond time scale of the sheath 110 formation and virtual cathode phase of ion acceleration). By the Boltzmann relation, the hot electron density, nhot, is related to the potential, φ, bynhot=n0 exp(eφ/kThot).  (Eq. 1)where n0 is hot electron density in the center of the target 104, k is Boltzmanns constant, and Thot is the temperature of the hot electrons. One may express the electric field Ex=−∂xφ, by the logarithmic derivative of nhot, that is,Ex=−kThot(nhot−1∂xnhot).  (Eq. 2)where ∂x is the hot electron scalelength. In the quasineutral regime, the electric field can be expressed by substituting nion˜nhot into Eq. 2. As long as the initial hot electron distribution is spatially smooth, radially symmetric, and either constant or monotonically decreasing with radius, the ion expansion exhibits laminar behavior, and the beam emittance is sufficiently low. The quasineutral plasma acceleration (including both accelerated ions/protons and electrons) decreases in time as Thot is reduced by the coupling of electron energy into the expanding ions and by radiative (i.e., Bremsstrahlung) and ionization processes in the bulk substrate material as the electrons oscillate through the bulk target material; and as the scale length of the ion (and hot electron) density increases due to the expansion.
It has been suggested that the entire beam envelope of the accelerated beam of protons and ions may be focused through the proper shaping of the rear surface of the target in order to produce a rasterable or scannable beam for lithographic and/or implantation applications. See Wilks et al., “Energetic Proton Generation in Ultra-Intense Laser-solid Interactions”, Phys. Plasmas, Vol. 8, pg. 542 (2001), which has been previously incorporated by reference herein. As theoretically illustrated in FIGS. 3A-3D, the non-irradiated surface 304 of the target foil 302 has been shaped to focus the entire beam envelope. As illustrated in FIG. 3A, a hemispherical portion 306 of the rear surface 304 of the target 302 has been removed, such that as illustrated theoretically in FIGS. 3B and 3C, that the entire beam envelope of the ion accelerated beam is focused, e.g., focused beam 308. FIG. 3D theoretically illustrates the ion beam intensity at the focus point for such a focused beam 308. Such a focused beam 308 would then be used as a raster to create a desired pattern in a sample material.
However, as illustrated in FIG. 4, this ballistic focusing of the entire beam envelope has so-far proven to be extremely difficult because one must compensate for the natural divergence of the beam envelope due to the electron sheath density gradients. In other words, the shape of the rear surface of the target 302 must compensate for the spatial distribution of the initial laser produced electron density 402 produced by the laser irradiation (i.e., the MeV electrons e− in FIG. 1 that produce the sheath 110). Sufficient control and reproducibility of the sheath distribution has not yet been achieved; thus, no one has been able to successfully focus the entire beam envelope as theoretically illustrated in FIGS. 3A-3D.
It is with respect to these and other background information factors that the present invention has evolved.